Background
HitTrax is a baseball simulator/launch angle monitor that measures real-time data, displaying live results for immediate feedback and analysis of key performance metrics. However, for personal consumer type use, HitTrax is expensive. The FlightsScope MEVO comes at a considerably lower cost but lacks a significant number of features the HitTrax offers.
We performed a simple data comparison to assess agreement between the FlightScope MEVO and HitTrax when hitting from a tee. The idea was to see if the FlightScope MEVO was a viable alternative for the personal consumer, Travel Teams, and High School Teams with a tight budget.
Assessing Agreement using Bland-Altman Plots [1,2,3]
The goal was to decide if the FlightScope MEVO agrees suitably with HitTrax, and hence whether the two can be used interchangeably for hitter assessments and to track hitters improvements due to a training program.
There are various methods for assessing agreement (interchangeability), however, the most popular and one of the most simple methods is limits of agreement approach (Bland-Altman Plots). The limits of agreement approach try to understand the observed differences between the measurements of two methods/devices by:
Plotting the two sets of measurements along with the line Y=X. If the measurements are comparable, they will be tightly scattered about the line.
Plot the difference between the measurements of the two devices against their mean. It is easier to judge departures from a horizontal line than from a tilted line. If the measurements are comparable, the differences should be small and show no systematic variation with the mean of the measurement pairs.
Construct confidence interval for the mean difference. This supplements plots with formal analysis.
Assuming no warning signs are raised by the plot in step (2), (that is, if the differences are small and there is no systematic variation of the difference with the mean) the measurements can be considered comparable (or interchangeable).
During this work we used a scatter plot of the observed difference between the measurements of the FlightScope MEVO and HitTrax (Difference=MEVO-HitTrax) against the average of the measurements of both devices (Average=(MEVO+HitTrax)/2). The purpose of this plot is to evaluate whether the differences are related to the averages, a surrogate for the unknown true values. If no relationship appears to exist, the distribution of the differences is summarized by the limits of agreement (LOA), which were defined by the average and standard deviation (LOA=Average+/-(1.96*StdDev)).
Equipment Used
Flightscope MEVO
HitTrax System
Set Up
FlightScope MEVO was placed 7 feet behind the hitter and down at the level of the ball on the tee. The HitTrax was placed 21 inches in front of the plate and 84 inches to the side of the cage (right batter box side).
Procedure
110 swings were taken of the tee (intended use of the FlightScope MEVO).
The FlightScope MEVO and the HitTrax were taking measurements of each swing.
The FlightScope MEVO was fed the classification of the result of the swing (groundball, flyball or line drive) based on the classification given by HitTrax (for comparison purposes).
The FlightScope Mevo missed the results of 23 swings, these swings were excluded from the analysis.
The Bland-Altman plots were created to measure/assess agreement between the FlightScope MEVO and HitTrax.
Results
The data below represent an attempt to determine whether the Exit Velocity, Launch Angle and Batted Ball Distance measurements made with a FlightScope MEVO could be used in place of HitTrax when tracking hitters improvement/progress due to a training program.
Exit Velocity
The plots of the FlightScope MEVO values against HitTrax values and their difference against their mean shows that the majority of the differences lie within the limits of agreement. Therefore, we can say that the exit velocity measurements from FlightScope MEVO and HitTrax are comparable or in agreement. In other words, when hitting from a tee the exit velocities measurements from the two devices could be used interchangeably to track improvements.
Launch Angle
When looking a launch angle the plots of FlightScope MEVO values against HitTrax values and their difference against their mean shows that for positive launch angles both devices were within 5 degrees of each other. However, for negative launch angles the difference between the two devices are within 20 and 25 degrees. Therefore, we can say that the launch angle measurements from FlightScope MEVO and HitTrax are only comparable or in agreement for positive launch angles.
Batted Ball Distance
Similar to the launch angle plots, the distance plots of FlightScope MEVO values against HitTrax values and their difference against their mean shows that only for distances greater than 125 feet both devices measurements were comparable. However, for distances below 125 feet the difference between the two were going outside the limits of agreements. These results make sense since the distance estimation for both devices depend on the launch angle measurements. Therefore, a disagreement in lower launch angles is more likely to affect the estimation of the shorter distances.
Conclusions
After looking at the data from FlightScope MEVO vs. HitTrax and their difference against their mean we observed that
The exit velocity measurements from FlightScope MEVO and HitTrax were comparable (in agreement).
The launch angle measurements from FlightScope MEVO and HitTrax are only comparable for positive launch angles.
The batted ball distance estimations were comparable (in agreement) only for distances greater than 125 feet.
The good news is that although the measurements from the FlightScope MEVO and HitTrax were not in agreement for the full range of the obtained data (launch angle & Distance) there were still ranges that had good agreement. The ranges were the launch angle and batted ball distance were comparable still allowed for the FlightScope MEVO to be used interchangeably with HitTrax for tee work when trying to get a rough picture of hitters improvements based on swing changes or a training program in combination with periodic HitTrax assessments performed in a facility or by an instructor.
References
Bland JM and Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; i: 307–310.
Brayne L, Barnes A, Heller B and Wheat J. Using a wireless consumer accelerometer to measure tibial acceleration during running: agreement with a skin-mounted sensor. Sports Engineering, December 2018, Volume 21, Issue 4, pp 487–491
Hanneman SK. Design, Analysis and Interpretation of Method-Comparison Studies. AACN Advanced Critical Care 19(2):223-34 · April 2008.
Kommentarer